雅思阅读 64: The Robotaxi Safety Debate（无人驾驶出租车的安全之争）

改编自 Traffic Injury Prevention / IIHS（2025-2026年）。雅思阅读 Section 3 难度，约 1050 词。 素材来源：https://arxiv.org/pdf/2505.01515v1

Reading Passage

A. For most of its short history, the autonomous vehicle industry asked the public to take a leap of faith. Promises that robotaxis would one day drive themselves were easy to make and impossible to verify, because every test car still carried a human safety driver ready to grab the wheel. That ambiguity ended between 2022 and 2025, when a handful of companies began operating vehicles with no one behind the steering wheel at all, picking up paying passengers on open city streets. The leader in this shift was Waymo, whose service spread from Phoenix into San Francisco, Los Angeles, Austin, Atlanta and Miami. By late 2025, the company had recorded more than 127 million fully driverless miles. The moment marked a genuine turning point: the question was no longer whether a car could drive itself in a controlled demonstration, but whether it could do so, every day, in traffic that is chaotic, unpredictable and overwhelmingly controlled by fallible human beings. Roughly 2,400 of the world's 3,500 commercial driverless robotaxis now belong to this single fleet, which means its record is watched not by enthusiasts alone but by regulators deciding whether the technology deserves to expand further.

B. The safety claims that followed were striking and, to many observers, surprising. A 2025 peer-reviewed study in the journal Traffic Injury Prevention analysed 56.7 million rider-only miles and compared Waymo's crash rate to human benchmarks on the same roads. It found a statistically significant reduction of about 85 percent in crashes causing suspected serious injury, and a sharp drop in injury-involving intersection collisions. An independent analysis later supported these figures: robots appeared to crash far less often than humans per mile driven. Reinsurance giant Swiss Re, whose business depends on pricing risk accurately, was brought in to validate the data externally. The numbers told a consistent story: a machine never gets drunk, never checks a phone, never falls asleep, and reacts within milliseconds to hazards that a distracted human might miss entirely. Proponents argue that if these trends held across millions of miles, robotaxis would not merely be novel — they would be among the safest vehicles ever to share the road. And because the system never tires or turns its attention elsewhere, a robotaxi accumulates, in a single year, the equivalent of many human lifetimes of driving experience.

C. Yet safety statistics do not, by themselves, settle a social question. Public trust has moved far more slowly than the data. Survey after survey has shown that a large share of drivers would still be afraid to ride without a human at the wheel, even when shown the crash figures. This reluctance is not irrational. Robotaxis, for all their strengths, encounter situations that human drivers resolve intuitively but that confound even the most advanced software: an emergency vehicle approaching with flashing lights, a child chasing a ball into the street, a construction crew holding vague hand signals. Every few months a video circulates online of a self-driving vehicle confused by roadworks, blocking traffic or failing to yield. A single awkward incident is viewed by millions, while a million uneventful miles are invisible. The result is a gap between actuarial safety and perceived safety that regulators now have to navigate. The data says the cars are safer; the public feels they are not. Trust surveys have barely moved even as the crash numbers improved, and this stubborn gap has become the central political problem of the industry, because no amount of statistical safety matters if voters refuse to share the road.

D. The comparison with human drivers also hides important complexities. Human-driven crash rates are themselves averages, drawn from every age, skill and state of attention behind the wheel. Robots operate only in carefully mapped suburbs and cities, in clear weather, at limited speeds — a narrow slice of the driving environment. Critics note that the most dangerous miles humans drive, on icy mountain roads at night, are precisely the miles robotaxis have not yet attempted. A machine that only ever drives on well-lit, well-mapped Phoenix boulevards may look flawless while being untested where danger actually concentrates. Insurance models and liability rules lag behind the technology: when a driverless car crashes, the question of who is responsible — the operator, the manufacturer, the software firm — remains unresolved in many courts. These are not objections to the statistics, but reminders that statistical safety in one operating domain does not automatically generalise to the whole road network. An analogy helps: a person who always walks on level, well-lit paths is statistically far less likely to fall than someone who climbs mountains, but this tells you nothing about how they would cope on a scree slope.

E. What happens next depends on whether the robots can keep their record as they expand. The coming years will test the fleet in rain, fog, snow and unfamiliar cities, and each new operating condition will reveal new edge cases that training data did not anticipate. Regulators, for their part, are now less willing to rely on companies' self-reported numbers and are pushing for independent, audited benchmarks that every manufacturer must meet. If the trend of fewer serious injuries holds up under those tougher conditions, robotaxis could gradually shift from an experiment in a handful of cities to a default mode of urban travel. If it does not — if the first dramatic, fatal crash erodes public confidence faster than the slow accumulation of safe miles can rebuild it — the technology may stall at the edge of the markets it entered. The arithmetic already favours the machines. The harder task is winning the argument with the human beings who share the road. That argument will be decided less by crash statistics than by how gracefully the robots handle the messy, improvisational moments — a cyclist swerving around a pothole, a delivery van double-parked in the rain — that make city driving such a peculiarly human art.

Questions 1-4

Choose the correct heading for paragraphs B, C, D and E from the list of headings below.

List of Headings i. The gap between statistics and public trust ii. How robotaxis first began to carry paying passengers iii. The impressive safety data and its validation iv. Why humans drive more dangerously v. Limits of the human-vs-robot comparison vi. The future test of expanding into harder conditions vii. The history of car insurance

1. Paragraph B: ____
2. Paragraph C: ____
3. Paragraph D: ____
4. Paragraph E: ____

Questions 5-8

Choose the correct letter, A, B, C or D.

5. What marked the turning point described in paragraph A? A. Robotaxis began to carry paying passengers with no human driver. B. The first robotaxi was sold to a private owner. C. Human drivers were banned from city centres. D. Robotaxis reached a top speed of 120 km/h.

6. What did the 2025 peer-reviewed study find? A. Robots crashed more often than humans. B. There was an 85 percent reduction in serious-injury crashes. C. The data could not be statistically analysed. D. Only Swiss Re trusted the figures.

7. Why does the writer say public distrust is "not irrational"? A. The statistics were later shown to be fabricated. B. Robots still face situations human drivers resolve intuitively. C. Human drivers are actually safer than robots. D. Robotaxis always refuse to move.

8. According to paragraph D, what weakness lies in comparing robots to human drivers? A. Human drivers are never tired. B. Robots only operate in a narrow, relatively safe set of conditions. C. Human crash rates are impossible to calculate. D. Robots drive at unlimited speeds.

Questions 9-13

Do the following statements agree with the claims of the writer?

Write:

• TRUE if the statement agrees with the information
• FALSE if the statement contradicts the information
• NOT GIVEN if there is no information on this

9. By late 2025, Waymo had recorded more than 127 million fully driverless miles.
10. Swiss Re was brought in to independently validate the safety data.
11. Most people surveyed said they would happily ride without a human driver.
12. Robotaxis already drive regularly on icy mountain roads at night.
13. Waymo is owned and controlled by a single car manufacturer.

Questions 14-15

Complete the summary below using NO MORE THAN TWO WORDS from the passage.

Although robots never get drunk or distracted, the public remains (14) ________ of riding in driverless cars, partly because each highly publicised (15) ________ erodes confidence faster than millions of safe miles can restore it.

答案与解析

← 上一篇 | 返回雅思焦点 | 下一篇 →